Skip to main content

Full text of "potts"

See other formats



Ariel M. Bagg 

1 Introduction 

The Ancient Near East is often referred to as the Fertile Crescent, a region of 
rich soils ranging from Palestine to Syria and Mesopotamia; in the south, at the 
crescent's concavity, it borders on the Syro-Arabian desert, in the north, at 
the crescent's convexity, it borders on the Anatolian and Iranian highlands. This 
image of a fertile, homogeneous region is a very simplified one, taking into 
account the great variety of ecological zones in this part of the world. In fact, 
landscape discontinuity is one of the structural features of the Middle East. Fur- 
thermore the term "fertile" has to be explained when applied to a region char- 
acterized by its aridity. It is generally accepted that a minimum of 200 millimeters 
of rain is necessary for rain-fed agriculture. Even if this is another simplification 
(see below), a look at mean annual rainfall in the Middle East shows that some 
two-thirds of the Fertile Crescent lie in the zone between 100 and 200 millim- 
eters, where only irrigation makes agriculture possible (Alex 1984). As a great 
part of the Ancient Near East was made fertile by man, namely by means of 
irrigation techniques, irrigation represents one of the most distinctive features 
of this cultural area. 

Most recent evidence on climatic development confirms the common assump- 
tion that during the last 6,000 years the climate of the Middle East widely cor- 
responded to today's conditions. Nevertheless, fluctuations of temperature and 
precipitation, differing in duration and amplitude, are attested, and their influence 

A Companion to the Archaeology of the Ancient Near East, First Edition. 

Edited by D.T. Potts. 

© 2012 Blackwell Publishing Ltd. Published 2012 by Blackwell Publishing Ltd. 

262 Developments in Farming, Animal Husbandry, and Technology 

at a local or more general level must not be forgotten. In historical times at least 
three first-order anomalies with a wide range of influence have been identified, 
the so-called "dry shifts" at around 3000, 2200, and 1300 BC. Other anomalies 
were weaker or affected smaller areas for shorter periods of time (Butzer 1995: 
134-8). Nowadays, the Fertile Crescent comprises mainly three climatic zones 
(Kottek et al. 2006): warm temperate climate with dry summer (the northern 
narrow fringe); steppe climate (the middle narrow fringe); and desert climate (the 
southern fringe). Characteristic of warm temperate and arid climates is the exist- 
ence of only two pronounced seasons, a dry and hot summer and a humid and 
cold winter. Rainfall is not the only, but one of the most critical factors that affect 
the growth of crops. The abovementioned minimum of 200 millimeters of rain 
may indicate whether or not dry-fanning is possible, but only if used correctly, 
taking into account additional aspects. 

The main problem lies in taldng the mean annual value of 200 millimeters and 
the corresponding isohyet as the limit between the two regions. It is not the 
annual average but the reliable annual aggregate that is to be considered 
for practical purposes. When the reliable rainfall amount, defined as the annual 
value which was reached or exceeded in 80 percent of the observation years, is 
considered, the position of the critical 100-300 millimeter isohyets changes 
dramatically, much reducing the dry-farming areas (Alex 1985). Thus, risldess 
dry-farming cannot be guaranteed in Iraq in areas receiving a below-average 
annual rainfall of 400 millimeters (Wirth 1962: 23). Furthermore, because of 
considerable, annual fluctuations in rainfall, it is not correct to conceive of the 
border between dry-fanning and irrigation agriculture as a line or even a strip 
- the 200 millimeter isohyets of a wet and a dry year may be up to 200 kilometers 
apart (Wirth 1962: 20-1). Therefore, between a risldess rainfall zone and a zone 
depending on irrigation, there exists no borderline, but rather an extensive belt, 
where dry-farming is possible but not assured (Oates and Oates 1976a: 111-13). 
In these areas of low to high risk, irrigation is necessary to ensure crop growth. 

Irrigation in the ancient Near East is frequently associated with water shortage, 
which is not always the problem. In some cases, the problem is having the right 
amount of water at the required time. Poor rainfall is compensated for in some 
areas, like southern Mesopotamia (Sumer and Babylonia), by plentiful surface 
water, which not only makes irrigation possible but also requires measures to 
prevent the injurious effects of a water surplus. In other areas, like northern 
Mesopotamia (Assyria), seemingly sufficient water resources cannot be exploited 
because of the great differences in elevation between the river and the fields, 
a situation that was only overcome by tapping and directing water sources 
from far away. Therefore, the two main rivers in the Middle East, the Tigris and 
the Euphrates, while closely related to irrigation agriculture, are part of both the 
problem and the solution. Both rivers follow a similar pattern: water levels 
increase over the winter months and reach a maximum in April or May at the 
time of the spring rains and the melting of snow in the highlands. Nevertheless, 

Irrigation 263 

the regime of the two rivers shows important differences. Both rivers rise on the 
Turkish plateau, but while the Euphrates, after its confluence with the Khabur 
river, runs for some 1,200 kilometers without being joined by any perennial 
tributaries, four main tributaries join the Tigris from the Zagros mountains to 
the east. These have steep slopes and carry erosion products. Heavy rains produce 
flash floods, which are common in the lower reaches of the Tigris but unknown 
on the Euphrates. The floods of the Tigris are unpredictable and potentially 
disastrous, whereas those of the Euphrates are not as violent and occur mostly 
in April or May. 

Cereals grow in the Middle East in winter. Sowing time is October to Novem- 
ber, while ripening takes place in April and May. The regime of the twin rivers 
does not at all fit the needs of agriculture (Ionides 1937: 4). For example, when 
irrigation water is most needed, river levels are at their lowest (September- 
October) and the spring flood coincides with the last phase of the growing period 
(April-May). A late spring flood shortly before the harvest can produce irremedi- 
able damage and even the loss of the crop. The problem is in this case not water 
shortage but, rather, the irregularity and unpredictability of the water supply. 

In Southern Mesopotamia, where both rivers flow with a gentle slope and tend 
to meander over the alluvial plain, another problem has to be considered. On 
the flood plains the rivers follow a meandering course with large marshes between 
them. Under these conditions both rivers, especially in the spring months, tend 
to change their courses, as attested by fossil meanders (Ionides 1937: 213-31). 
In the delta plains, where the slope is reduced to only 3 centimeters per kilometer, 
the rivers tend to split into a number of branches. Their course in the 2nd and 
1st millennia BC is still unclear in many points, but some reliable reconstructions 
have been proposed (Cole and Gasche 1998; Gasche et al. 2002). Furthermore, 
both rivers carry great quantities of sediments which are deposited during the 
flooding. The deposited silt creates natural levees and raises the riverbed, causing 
the river to flow above the plain level. Some sediment is deposited in the canals 
and in the riverbeds, so that these need regular cleaning to enable water to con- 
tinue to flow in the desired amount and direction. Nevertheless, much of the 
sediment ends up in the fields, with a negative effect on the soil. Fine sediments 
settle on the soil surface or may move into deeper soil layers, hindering water 
infiltration and the emergence of seedlings. 

Irrigation is essential to make agriculture possible beyond the dry-fanning 
regions, to enhance productivity, and to enable more than one crop per year. 
Nevertheless, some side-effects of irrigation have a negative impact and must be 
taken into account when planning irrigation systems. The combined and related 
effects of excessive salt-accumulation in the root-zone and the development of a 
high water table are the basic causes of crop failure under irrigation. Salt comes 
from irrigation or floodwater, and most of the soils in southern Iraq are saline 
to some degree. In arid climates, where evaporation exceeds precipitation and 
the water table rises because of the surplus irrigation water, salts cannot be 

264 Developments in Farming, Animal Husbandry, and Technology 

leached out of the topsoil into the subsoil. When the ground water reaches depths 
of 1-2 meters below the surface and comes within reach of the evaporative force, 
water is lost to the atmosphere leaving the salts in the upper soil layers. Not all 
salts are harmful to crops, but some (chlorides and carbonates) are toxic. Saliniza- 
tion can affect crops in different ways. Plants may struggle to obtain certain key 
nutrients, but the main problem is the increase in concentration of the soil solu- 
tion, increasing the pressure that plants need to apply in order to extract water 
from the soil. In conditions of highly saline situations, plants may suffer physical 
or physiological damage. Concentrations of salt from 0.1-0.2 percent begin to 
be injurious to crops and concentrations of 0.5-1.0 percent become intolerable 
(Oates and Oates 1976a: 124). The application of surplus irrigation water is a 
relatively simple method of washing salts out of the rooting zone. Fallow rota- 
tion, attested also in historical times, is another practice intended to alleviate the 
effects of salinization. 

In order to overcome the difference in elevation between the level of the water 
and that of the surface where the water is needed, water- lifting devices are neces- 
sary. Of the different methods for lifting water from rivers, canals, pools, or wells 
used in preindustrial societies (Molenaar 1956), only the pulley, the shaduf {sec 
below) and a land of chain of pots are clearly attested in ancient Near Eastern 
written and iconographic sources (Bagg 2001: 41-4). The existence of other 
devices, like animal -powered Persian wheels (sakiya), rope -and- bucket lifts (cerd), 
man-powered paddle-wheels, water-wheels, and Archimedean screws, has been 
suggested, but such proposals are based on obsolete readings or controversial 
interpretations of the cuneiform sources (Bagg 2001: 44-6; Volk 2009). Up to 
now, solid written or archaeological evidence for these water-lifting devices is 

The counterpoise lift, known as shaduf (one of its many Arabic names), is an 
easy-to-construct but highly efficient device for raising water. The working prin- 
ciple is very simple. A long, wooden pole is pivoted as a lever from a crossbar 
supported by one or two pillars. A large stone or a clump of dried mud is fixed 
to the shorter end of the lever, serving as a counterpoise to a bucket-type dipper 
suspended from a rope or rod attached to the longer arm of the lever. A man 
needs only to pull down on the rope or rod until the container enters the water 
and fills up. Then, he allows the lever to lift the full bucket to the required height 
at which point he empties it by tipping it sideways. The shaduf has a working 
range of 1-3 meters. In case of lifts exceeding 3 meters, two or more devices can 
be used in series. The performance of one man, considered as the average rate 
of raising water during a full working day, is 3 square meters of water per hour 
(Molenaar 1956: 8). 

The earliest depiction of a shaduf in ancient Mesopotamian art appears on a 
cylinder seal from the late Akkadian period (c.2200bc) (Figure 14.1). Older 
written attestations (c. 2450-2350 BC) are found in pre-Sargonic texts from Girsu 
(Bagg 2001: 40-1). The use of shadufs in Assyria (northern Iraq) in the 1st mil- 



Figure 14.1 Cylinder seal showing the use of the 
Fig. 397, Louvre A. 156). 

C.2200BC (afterWard 1910: 

Figure 14.2 Two-stage shaduf installation (7th century BC). Detail from a relief in 
Sennacherib's Southwest Palace at Nineveh (BM 124820) (after Davies 1933: Fig. 10). 

lennium BC is documented on a relief from Sennacherib's (704-681 BC) palace 
at Nineveh (Figure 14.2), belonging to a cycle hi which the transport of one or 
more bull colossi is shown. In the lower part of the scene a two-stage water-lifting 
installation is depicted, consisting of three levers of two (one of them double) 
or three (simple) shadufs, which are operated in each case by one man standing 
on a platform. The pillars are made of masonry and their height corresponds to 
that of a man. The levers are approximately 3 meters long and probably rest on 
a wooden beam (not shown). At their ends are counterweights. The conical 
buckets used were likely made of leather. The platform on which the men stand 
may be understood as a canal parapet or a basin wall (Bagg 2000: 204-7). 

266 Developments in Farming, Animal Husbandry, and Technology 

The second water-lifting device attested is the pulley used for lifting water from 
wells, used in the ancient Near East to irrigate gardens. A bucket is attached to the 
end of a rope, which passes over a pulley set in a framework over the well. 
The earliest depiction dates to the 9th century BC and is found on a wall relief 
(now in the British Museum: BM 118906) from Assurnasirpal IPs (883-859 BC) 
palace at Kalhu (modern Nimrud). Over the city wall of a besieged city a pulley 
can be seen, and outside the city a soldier is shown cutting a rope from which a 
bucket is hanging. Pulleys dating from the 8th-7th centuries BC were found at 
Dur-Sharrukin (modern Khorsabad) and Nineveh (Bagg 2000: 105-6, with Pis. 
18, 21a). Also related to wells and probably used for the irrigation of palace 
gardens is a third device attested in one of Sennacherib's inscriptions. The Icing 
speaks proudly of a technical innovation for raising water that replaced the 
common shaduf. We know only that the device consisted of pulleys, bronze chains, 
and bronze wires, and that it was positioned over a well by means of a metal 
support. Maybe it was a land of chain of buckets or pots (Bagg 2000: 199-203). 
Subterranean galleries to tap groundwater for irrigation purposes (Persian qanat), 
as attested in Iran from the Hellenistic period onward, are, contrary to common 
opinion, not attested in the ancient Near East (Bagg 2000: 12746; Salvini 2001). 

Before discussing some paradigmatic cases of ancient Near Eastern irrigation 
practices in regions where they are archaeologically well documented, some 
terminological comments may be useful. Irrigation is defined as the artificial 
application of water to the soil. When agriculture can be practiced relying only 
on rainfall this is called "dry-farming." When the minimum amount of rainfall is 
unreliable or does not reach the quantity required for dry-farming, the soil needs 
to be artificially moistened, referred to as "irrigation agriculture." There exist 
different types of irrigation techniques that vary in the way in which the water is 
distributed within a field. Characteristic of the ancient Near East is surface irriga- 
tion, by which the water is distributed along the field using gravity flow (Booher 
1974). In the ancient Near East the distribution always happened by means 
of open distribution systems (as opposed to piped distribution systems) - namely, 
open canals located at the high edge of the field from which the water could be 
directed into basins or furrows. Two main types of surface irrigation techniques 
were in use in the ancient Near East: basin irrigation and furrow irrigation. 

Basin irrigation was the simplest way to irrigate fields and was therefore widely 
used. In this system, fields are divided into units with a nearly level surface. Levees 
(earth banks) are constructed around the fields, forming basins. The water is 
directed into the basins up to the desired depth and retained until it infiltrates 
the soil. Eventually, any excess water can be drained off. Variations of this method 
relate to the size and shape of the basins, the techniques for directing the water, 
and continuous or rather intermittent ponding. 

Furrow irrigation consists in letting water run in small channels (furrows) that 
carry the water as it moves down the predominant slope of the field. The water, 
applied to the top end of the furrows, sweeps into the bottom and sides of them, 

Irrigation 267 

providing the required moisture. In contrast to basin irrigation, the entire soil 
surface is not moistened. This method is suitable for the irrigation of orchards 
and vineyards, as well as crops which could be harmed if the water were to reach 
the top or the stems of the plants. 

An irrigation system is a network of channels and control structures in a cul- 
tivable area used to transport water from its source (a river, a main reservoir) to 
the fields. A canal is a manmade channel or canalized, natural watercourse which 
forms part of an irrigation system. Within an irrigation system there are different 
categories of canals according to their dimensions or, in other words, to the 
amount of water they transport, which follow a hierarchical pattern. Main canals 
transport water from the source to a secondary or branch canal and have, there- 
fore, a major cross-section. As the channels approach the crops, their cross-section 
diminishes, and there are then secondary, tertiary, quaternary, etc. channels. The 
structures needed to distribute and control water are diversion works located at 
the head of the system which allow water to be diverted from the source to the 
system. Regulators are structures across the channels to maintain water levels and 
to control the water supply. 

2 Southern Mesopotamia (Sumer and Babylonia) 

Agriculture in southern Mesopotamia was impossible without irrigation. The 
region not only lies beyond the limit of dry-farming, but the Tigris and Euphrates 
are at their lowest when water for the crops is most needed and flood at harvest 
time. Both rivers derive their waters from winter snows in eastern Turkey. The 
Euphrates has a slower flow and its bed lies over the surrounding plain, allowing 
floodwater to fill the adjacent basins and to stay there. Therefore, the Euphrates 
is more suitable for irrigation purposes than the Tigris, which has a lower bed 
and a more violent and unpredictable flood. In order to cultivate the land, it was 
necessary in ancient times to control and direct the floodwaters through a 
network of canals, which started at the levees and conveyed the water to the field 
by means of smaller canals following a dendritic pattern. Work invested in canal 
maintenance and in protective works must have been at least as important as the 
work demanded for digging the network. Sumerian and Babylonian texts contain 
much information on these activities (e.g., Stol 1988; van Driel 1988; Renger 
1990; Waetzoldt 1990). 

In the late 4th and 3rd millennia BC there was a concentration of settlements 
in the delta region, where cities like Ur, Uruk, Umma, and Larsa were located. 
There the floods were generally not as violent as in the northern river plain. With 
the technology available, it would have been more difficult to construct and 
maintain a network of irrigation canals in the north than in the delta region, 
which may in part explain this settlement pattern (Butzer 1995: 142-5). The 
importance of irrigation in southern Mesopotamia, as well as the fact that some 

268 Developments in Farming, Animal Husbandry, and Technology 

degree of social organization was needed to create and maintain canal systems, 
is undoubted. In the past, this view led to the so-called "hydraulic hypothesis," 
which explained the increasing complexity of societies - e.g., in Mesopotamia - 
and therefore the rise of urban culture, as the result of the need to construct and 
administer irrigation works (Wittfogel 1957). This view is no longer compatible 
with our knowledge of the facts, mainly because the existence of complex irriga- 
tion systems in the early periods (7th to early 4th millennium) can be ruled out, 
as we know from (Late) Babylonian sources (Nissen 1988: 58-60). 

We do not know when irrigation was first practiced in southern Mesopotamia, 
but the earliest archaeological evidence dates from the early 6th millennium and 
comes from Choga Mami, near Mandali, a site in the foothills of the Zagros 
mountains. In modern times the palm gardens of Mandali were irrigated by 
means of a fan of channels fed from the river Gangir, a method that also seems 
to have been applied in ancient times. On the northern slope of the mound, 
seven small, manmade water channels (each c.2 meters wide), dating to shortly 
after 6000 BC, were discovered. On the southern slope of the site a major canal 
(c.10 meters wide), filled with pottery of the Samarra period (late 6th millen- 
nium), was excavated. Furthermore, superficial traces of this and another canal 
were discovered in the vicinity of the site, suggesting continuous use through 
the Ubaid and possibly the Uruk periods (Oates and Oates 1976a: 128-33). 

Even if irrigation cannot be considered the prime motor of complex societies, 
it demands, in fact, more physical and intellectual energy than dry-farming. Irri- 
gated fields can be farmed more intensively and greater yields can be achieved. 
As a consequence, more people can be fed from the yield of a given area; in other 
words, a smaller irrigated territory around a settlement will feed the same popula- 
tion as a larger, unirrigated area, so that the settlements could lie closer to each 
other than in dry-farming regions. Indeed, an increase in settlement size and 
density is attested in the Mesopotamian floodplain soon after the beginning of 
the 4th millennium BC, as is the systematic use of irrigation from the Late Uruk 
period (late 4th millennium BC) onward. Since then, the lower Mesopotamian 
plains have been intensively farmed and irrigated by a canal network right up to 
the present day. Dating canals is an extremely difficult task, because unless they 
are clearly associated with settlements which flourished in a certain period, only 
a probable chronological assignment can be postulated. Earlier canals were often 
reused and objects found in a canal bed give only a terminus post auem for their 
use. Nevertheless, in recent years a great deal of work has been done combining 
data from archaeological surveys, aerial and satellite imagery, and written sources. 
Maps showing settlements, canal remains, and reconstructed canal courses are 
available for the Diyala region in the 4th and 3rd millennia BC (Adams 1965: 
Fig. 2) and for the southern Mesopotamian plain in the 3rd millennium (Adams 
and Nissen 1972: 36, Fig. 17; Adams 1981: Figs. 29-31; Steinkeller 2001: 40, 
map 1). More accurate results using modern mapping technologies are available 
for the region between Ramadi and Babylon in the 2nd and 1st millennium BC 
(Cole and Gasche 1998: 49, maps 8, 9, and 51). 

Irrigation 269 

The principal purpose of the irrigation network was to supply the grain fields 
with sufficient water at the right time, a purpose connected with the solution of 
certain problems (Postgate 1992: 176-83) including (1) supply: water had to be 
directed onto the fields (canals, regulators); (2) storage: water had to irrigate the 
fields at the right time and as long as necessary (reservoirs); (3) drainage: fields 
had to be drained if water was no longer necessary in order to avoid damage 
(e.g., salinization); and (4) protection: fields had to be protected against water 
surplus (levees). The cuneiform texts show that the Sumerians were well aware 
of these technical problems. The rulers of the 1st Dynasty of Lagash (c.2600- 
2450 BC) mention in their building and votive inscriptions irrigation works, in 
particular the construction or repair of canals (Laurito and Pers 2002). A great 
deal of manpower was expended for the construction of permanent works made 
of bricks and bitumen, which served to convey water from a main canal. The 
regulation of the flow occurred most probably by means of removable wooden 
beams (Sum. ges-kes-du), built for instance by Enmetena and Uruinimgina. An 
impressive structure excavated at Telloh (ancient Girsu) may represent one such 
regulator (Parrot 1948: 211-19; Dight 2002), and inscriptions of Icing Pirigme 
(c.2200-2150 BC) found in situ mention that he built one (Edzard 1997: 12-13). 
From this period we have also the first report of conflicts surrounding the use 
of irrigation water, in particular the long-standing conflict between the cities of 
Lagash and Umma concerning the water supply for the Gu'edena, a fertile area 
of cultivation located between them (Cooper 1983a). 

The construction and ongoing maintenance of hydraulic works required for- 
midable labor organization. As shown by numerous economic and administrative 
documents of Ur III date (2100-2000 BC) - e.g., from Umma (Tell Jokha) - this 
task was undertaken by the state. Canals, levees, reservoirs, and water outlets were 
examined and measured to identify damage or blockages. Information about the 
number of workers, the duration and the land of the maintenance work is also 
documented in the texts (Sauren 1966; Waetzoldt 1990). A Sumerian literary 
work dating from the 18th or 17th century BC contains instructions from a farmer 
to his son and describes the tasks to be performed throughout the agricultural 
year (Civil 1994). Before the first irrigation by flooding, which took place at the 
time of the spring high water (April-May), the dykes and irrigation channels were 
to be thoroughly checked. Thereafter, the cultivated plants were watered four 
times during their growth cycle. 

According to a widespread view, increasing soil salinization should have led to 
a decrease of productivity and the decline of the Sumerian culture by the late 3rd 
or early 2nd millennium BC (Jacobsen and Adams 1958; Jacobsen 1982). Even 
if this theory has been shown to be incorrect (Powell 1985), salinization was a 
real problem at that time. The measures undertaken against it cannot be clearly 
identified in the texts, but fallow was apparently carried out as well as the leach- 
ing of the soil (Powell 1985: 36-8). 

The tradition of irrigation agriculture initiated by the Sumerians was continued 
from the 2nd millennium onwards in Babylonia. Cuneiform texts from the Old, 

270 Developments in Farming, Animal Husbandry, and Technology 

Middle, and Late Babylonian periods deal with the construction and maintenance 
of the canal network as well as with the accumulation of silt in canal beds and 
the use and regulation of irrigation water. Economic and administrative docu- 
ments, as well as letters from Larsa (Tell Senkereh) dating to the early 19th 
century BC, contain valuable information about irrigation there (Walters 1970). 
They refer not only to the excavation of canals, but also to surveys, silting, and 
the organization of manpower. It is not always possible to decide whether the 
information concerns the excavation of a new canal or the dredging of a silted -up 
canal, because the terminology is undifferentiated. Royal inscriptions mention 
canalization work on the Tigris undertaken by Sin-iddinam (Frayne 1990: 160, 
11.39-70) and the excavation of the Mami-sharrat canal by Rim-Sin (Frayne 1990: 
291-3). The rulers of the 1st Dynasty of Babylon, of whom Hammurabi is the 
most famous, also dealt intensively with irrigation works (Renger 1990). Legal 
and administrative documents, and especially letters, give detailed information 
about the problems to be solved. In some cases, Hammurabi or one of his officials 
gave instructions about the measures to be taken when either too little or too 
much water was available for irrigation (Kraus 1968: Nos. 13, 18, 19, 39, 74, 
80, 85, 109, 114, 131). A rich technical terminology was used in these texts for 
different types of canals, weirs, maintenance work (Stol 1976-80), and different 
fields (Stol 1988). 

From the Late Babylonian and Persian periods come thousands of texts with 
information about the organization of agriculture. Relevant legal and administra- 
tive documents have survived from archives of the Ebabbar (Shamash temple) at 
Sippar (modern Tell Abu Habbah; Jursa 1995) and the Eanna (Ishtar temple) 
at Uruk (modern Warka; Cocquerillat 1968). In addition, there are private 
archives of entrepreneurial families, such as the Egibi of Babylon (Wunsch 2000) 
and the Murashu of Nippur (Stolper 1985). These texts from temple and private 
archives deal primarily with lease contracts and agriculture personnel and deliver- 
ies, and less so with irrigation itself. Nevertheless, a rich technical vocabulary and 
numerous canal names are attested (Zadok 1985; van Driel 1988). The royal 
administration was responsible for the supervision of the canal system and the 
main irrigation projects, even if the work was carried out by local institutions. 
Tenants were responsible for the excavation and maintenance of the smaller canals 
that irrigated their own fields. This included oversight of canals, reservoirs, and 
dams in order to avoid flood damage. 

3 The Middle Euphrates 

The Middle Euphrates valley between the Syrian cities of Abu Kemal and Deir 
ez-Zor is an arid region with less than 150 millimeters of annual, highly irregular 
rainfall and fewer than 40 days of rain per year. The dry season is long, tempera- 
tures are high, the groundwater has a high concentration of chlorides and sulfates, 

Irrigation 271 

and arid winds (khamsin) blow in spring and summer. Under such conditions, 
dry-farming is impossible and agriculture requires irrigation. The spring flood 
(March-May) is violent and irregular, and the flash floods of the affluent wadis 
occur earlier (February-March) and are even more destructive. The river mean- 
ders at the bottom of the valley, about 30-40 meters below the level of the 
plateau. Nowadays, the Euphrates flows on a terrace formed in Roman and 
Islamic times (an "historical terrace"), some 2 meters below the level of the 
Holocene terrace where most archaeological sites are located. The width of 
the valley varies from 6 to 14 kilometers on the right bank but is much narrower 
on the left bank. The widenings and narrowings of the valley form three main 
basins, known in the literature as alveoli; the northern one extends from Deir 
ez-Zor to Bouqras, the middle one from Bouqras to Dura Europos, and the 
southern one from Dura Europos to Abu Kemal (Geyer 1990b: 63-6). In the 
last mentioned region lies Mari (modern Tell Hariri), which played an important 
role from the 3rd millennium to its destruction by Hammurabi in the middle of 
the 18th century BC. 

The Middle Euphrates valley has been extensively surveyed by archaeologists 
and geographers, and many studies have been devoted to the relationship between 
Mari and its environment (Geyer and Monchambert 1987). One of the results 
of these surveys was the discovery of ancient canal remains. In addition, agricul- 
ture and irrigation are well represented in the letters found in the palace archive 
of Zimri-Lim, the last king of Mari. The official correspondence deals with the 
irrigation of the administrative units of Mari, Terqa, and Saggaratum. As well as 
the repeated complaints about a shortage of workers, the letters inform us about 
different aspects of canal maintenance, the protective system of weirs and dams, 
as well as the extent of water damage, using specific technical terms (Durand 
1990, 1998: 573-653; Lafont 2000). Three main canals are mentioned in the 
sources: the Ishim-Yahdun-Lim canal (on the right bank of the Euphrates), 
the Mari canal (also on the right bank), and the Khabur canal (on the left bank). 
Fields up to 1 kilometer away from the river could be watered directly by drawing 
water from the Euphrates with simple water-lifting devices. In order to irrigate 
larger areas further from the river, however, an irrigation system was necessary, 
involving canals fed by the Euphrates or the Khabur. In fact, the different parts 
of the valley were referred to by specific terms, and a distinction was made 
between "fields irrigated by means of water lifting" (daluwatum) and "fields 
irrigated by means of canal water" (masqitum). 

With respect to the amount and quality of the available archaeological and 
written sources, the Middle Euphrates valley seems to be an ideal case for the 
reconstruction of ancient irrigation. Nevertheless, correlating data from archaeo- 
logical sources and texts is both difficult and contentious. On the one hand, it 
is not easy to date canals, as they may have been reused over time (Monchambert 
1987). On the other, the information in the sources, however rich, is restricted 
to a short period of time, lasting only a few decades, within the many centuries 

272 Developments in Farming, Animal Husbandry, and Technology 

during which the kingdom of Mari existed. On the right bank of the Euphrates 
the remains of a main canal were discovered, the course of which could be fol- 
lowed for more than 17 kilometers. The longest stretch, more than 1 kilometer, 
was discovered some 6 kilometers north of Mari. The canal bed is c.20 meters 
wide and has impressive dykes extending over almost 100 meters. The canal was 
certainly used for irrigation, and traces of some minor off-take canals were also 
found. This is the most probable candidate for the Mari canal mentioned in the 
texts. As agriculture at Mari was impossible without irrigation, the canal has been 
dated by the excavator to the 3rd millennium BC, contemporary with Mali's 
foundation (Margueron 2000: 75-9). 

A second irrigation canal, on the right bank of the Euphrates south of Deir 
ez-Zor, could be followed over a distance of 30 kilometers. The remains of three 
secondary canals branching off from die left bank of the canal were also discov- 
ered. Known in the Islamic period as Nahr Sa'id, this canal was probably in use 
in the Bronze Age, and is a good candidate for the Ishim-Yahdun-Lim canal 
mentioned in the Mari texts as flowing from the city of Dur-Yahdun-Lim (Deir 
ez-Zor?) to Terqa (Tell Ashara). Another main canal, the Nahr Daurin, is located 
on the left bank of the river and was at least 110 kilometers long. It has been 
suggested that it was already being used for navigation in the Bronze Age. 
However, both its function and dating are controversial. Finally, another, shorter 
canal ran directiy to Mari from the Euphrates and seems to have supplied the 
city with water. 

The Lower Khabur valley, to the northeast of the kingdom of Mari, was also 
intensively surveyed from an archaeological and geomorphological point of view 
in the late 1970s, by a German team. The Khabur is the main tributary of 
the Middle Euphrates. The region around the Assyrian provincial center 
Dur-Katlimmu (modern Tell Sheikh Hammad) lies between the 100 and 200 
millimeter isohyets - i.e., in a risk zone where crops can only be guaranteed by 
means of irrigation. In fact, a late Middle Assyrian text (10th century BC) 
mentions the repair of a canal in the Khabur region. Further information about 
irrigation is scarce, but Middle Assyrian letters from Tell Sheikh Hammad and a 
Neo-Assyrian inscription of Tukulti-Ninurta II (890-884 BC) mention irrigated 
fields and a canal related to the Khabur (Bagg 2000: 56-9). Ancient canals have 
been discovered and carefully mapped on both banks of the Khabur (Botsch 
1986; Ergenzinger and Kuhne 1991; Stellmacher 1991). They are 7 meters wide 
at the base, 8.5 meters wide at the water level, 1-1.5 meters deep and could be 
followed over a distance of 250 kilometers. The average slope is 0.03 percent. 
The western canal (called Nahr Ham'a) was fed by the Khabur, whereas the 
eastern canal (Nahr Daurin) was fed by the Wadi Jagjag. 

Most probably these canals were multifunctional, serving for irrigation, naviga- 
tion, the regulation of the Khabur in the flood season, and water supply (Morandi 
Bonacossi 1996: 97-9). As in the case of the Middle Euphrates canal system, the 
dating of the Khabur canals is problematic. The excavator dates the eastern canal, 

Irrigation 273 

which terminated at Dur-Katlimmu, to the Middle Assyrian period (13th century 
BC) and the western canal to the Neo-Assyrian period (9th-7th century BC; 
Ergenzinger and Kiihne 1991). Thereafter, both were in use from the Hellenistic 
to the Islamic period. However, considering the setdement pattern of the Middle 
Assyrian period, a regional canal system at this time seems improbable (Morandi 
Bonacossi 1996: 100-1). On the contrary, considering that Dur-Katlimmu grew 
from 15 to 100/120 hectares in the Neo-Assyrian period, the exigencies of 
feeding a larger population makes the construction of both canals in the 8th or 
7th century BC more probable. With the addition of a regional irrigation system, 
the agricultural potential of the Lower Khabur valley must have been very high, 
and a population of 30-45,000 is reasonable to suggest (Morandi Bonacossi 
1996: 194-204). The navigability of the canals has been studied and proven 
(Botsch 1986: 74-86; Ergenzinger and Kiihne 1991: 175). Navigation in the 
Lower Khabur valley is not attested in the written sources, but reference to a 
navigable "canal of Suhu" on the Middle Euphrates in an 8th century BC inscrip- 
tion (Bagg 2000: 58-9) makes the idea of navigable canals in the Lower Khabur 
valley plausible. 

4 Northern Mesopotamia (Assyria) 

Because of an over-simplified opposition - "irrigation agriculture in Babylonia/ 
dry-farming in Assyria" - the achievements of Assyrian hydraulic engineers in the 
field of irrigation were long overshadowed by those of their southern neighbors 
and even misunderstood as luxury works for watering royal gardens. However, 
as noted above, the border between the dry-farming zone and areas in which 
irrigation is necessary is not a clear line, but rather a transitional zone about 400 
kilometers wide between the 100 and the 400 millimeter isohyets. Rainfall varies 
from year to year. Rain falls from December to March, often heavily, but with 
strong variations in both geographical distribution and amount. All these varia- 
bles have a dramatic effect on the success or failure of the harvest. Considering 
that climatic conditions have not changed much during the past 6,000 years, it 
is clear that dry-farming was not possible in Assyria - particularly in the south 
- without high risk. Irrigation was necessary to guarantee crops and to raise yields. 
This is why the Assyrian ldngs carried out irrigation projects near their capitals, 
where a large population had to be nourished. As the diversion of water from 
the Tigris was difficult because of the difference in elevation between the river 
and the fields (up to 7 meters), water was instead conveyed over relatively long 
distances to the capital cities by means of main canals, mainly from the Greater 
Zab, the Khosr, the Atrush, and the Wadi Bastura. The Assyrians investigated 
the water resources of the neighboring mountain regions, diverting water from 
mountain streams and springs into canals. Wadis were even canalized and inte- 
grated into the canal system. The water was conveyed not only by artificial canals 

274 Developments in Farming, Animal Husbandry, and Technology 

to the cities. Low-volume rivers, like the Atrush and the Khosr, were also used 
as canals, fed with additional water. 

In the 14th century BC Assyria was a small kingdom in northern Iraq. After 
several phases of expansion, by the 7th century BC the Assyrian empire encom- 
passed the entire Middle East from Iran to Egypt. The cities which functioned 
as capitals all lay near the Tigris in Assyria's heardand, the borders of which were 
the Zagros mountains to the north and northeast, the Lesser Zab river to the 
southeast, Jabal Makhul in the southwest, and the Wadi Tharthar in the north- 
west. Assyrian irrigation projects are attested in written sources from the twelfth 
to die 7th century BC. With the help of cuneiform sources, iconographic material, 
and archaeological remains, it is possible to reconstruct the history of irrigation 
in Assyria and to understand the relevant technical terminology (Bagg 2000). 

The foundation of a new capital or the enlargement of an existing city to turn 
it into a capital is well documented in Assyrian history. The old commercial and 
cultic center of Assur (Qalat Sherqat) on the west bank of the Tigris was the 
capital in the Middle Assyrian period (14th-llth century BC) and the most 
important cultic center during the whole of Assyrian history. However, with an 
area of 70 hectares, it remained the smallest of all Assyrian capitals. The first of 
a series of new foundations was undertaken by king Tukulti-Ninurta I (1243- 
1207 BC), who established a royal residence on the east bank of the Tigirs, only 
3 kilometers upstream from Assur, and called it Kar- Tukulti-Ninurta - i.e., 
"Tukulti-Ninurta' s Harbor" (modern Tulul Al 'Aqar). According to cuneiform 
sources, the king looked for additional water resources in the mountains and 
directed spring water to the town to convert an uninhabited area into irrigated 
fields. The new city was also inhabited in the Neo -Assyrian period (10th-7th 
century BC) and was at least three times larger than Assur. It is clear that the 
Kar-Tukulti-Ninurta was planned for a large population and that irrigation was 
an important factor which had to be considered from the outset. The remains of 
a canal which flowed through the city were already discovered during the first 
excavations, as well as another canal, which came off the Tigris to the north of 
the city. Further canals were found in the Makhmur plain to the west of Kar- 
Tukulti-Ninurta. The dates of these canals are unknown and they may have been 
in use in later periods. 

The first of three cities which successively became imperial capitals in the 
Neo-Assyrian period was Kalhu (Nimrud) on the east bank of the Tigris, about 
8 kilometers upstream from its junction with the Greater Zab. Assurnasirpal II 
(884-859 BC) built there for 15 years and turned the city into a new royal resi- 
dence extending over 360 hectares. In his inscriptions the king wrote that he 
dug a canal from the Greater Zab called "Canal of Abundance." His purpose 
was to irrigate the fields and gardens in the flood plain of the Tigris. Assurnasir- 
pal II also described in detail a watered "pleasure garden" planted with exotic 
trees (Grayson 1991a: 290, 11.36-52). It is not known how much land was irri- 
gated, but estimates range as high as 2500 hectares. Assurnasirpal II's "Canal of 

Irrigation 275 

Abundance" was already identified in the 19th century and can be traced as a 
rock-cut channel along the right bank of the Greater Zab for some 8 kilometers 
from the village of Quwair to a point about 5 kilometers before the river joins 
the Tigris, where it flows in a northwest-southwest direction, to the southeast- 
ern corner of Kalhu. During part of its existence, the canal was fed by a rock-cut 
tunnel, the so-called Nagub tunnel, which passes through a conglomerate bluff 
on the right bank of the Greater Zab. An inscription found in situ mentions 
restoration works carried out 200 years later by the Assyrian Icing Esarhaddon 
(680-669 BC). Although badly damaged, the inscription refers to the repair of 
Assurnasirpal IPs canal, which no longer functioned because of an accumulation 
of sediments. 

Kalhu remained the Assyrian capital until Sargon II (721-705 BC) decided to 
found a new royal residence covering 300 hectares, about 50 kilometers away, 
called Dur-Sharrukin "Sargon's Fortress" (Khorsabad). The city was built over 
the course of 12 years and was probably never finished, as the Icing died unex- 
pectedly on the battlefield. Sargon presented himself as someone interested in 
land reclamation, the planting of orchards, the search for additional water sources 
in the mountains, and land irrigation. One of the reasons given for the construc- 
tion of the new city was to provide Assyria with abundant food. The texts mention 
the construction of a canal, but there is no evidence that it was ever built. In the 
tradition of importing exotic botanical specimens for their acclimatization in 
Assyria (for which purpose irrigation was needed in more than one case), initiated 
by Tiglath-pileser I (11 14-1076 BC), Sargon planted a new type of royal garden 
characterized not only by its exotic flora, but by a newly created landscape of 
ponds and artificial hills with pavilions on top. This park is depicted in reliefs 
from Sargon's palace (Bagg 2000: 156-9, Pis. 32-36). We do not know if the 
park was watered, but similar gardens created by Sargon's successors were cer- 
tainly irrigated, as shown below. 

Sargon's son Sennacherib also decided to change his residence and enlarged 
the old city of Nineveh, located where the Khosr joins the Tigris. He surrounded 
it with a 12 kilometer-long city wall and made it into the most splendid of all 
the Assyrian capitals, covering an area of 750 hectares. Po supply the new capital 
and the surrounding fields with water, Sennacherib undertook the most ambi- 
tious hydraulic project in Assyrian history: four canal systems, altogether more 
than 150 kilometers long, with canals and canalized watercourses, tunnels, aque- 
ducts, and weirs. Sennacherib's inscriptions enable us to follow the realization of 
this project between about 702 and 688 BC. In addition, the archaeological 
remains detected through surveys and satellite imagery have been essential in 
reconstructing this system (Ur 2005). Po supply Nineveh with water, new 
resources were tapped between the city and the eastern mountains and directed 
by four canal systems which reached the town from different directions following 
a radial pattern. The principal purpose of this enterprise was to increase the size 
of the cultivated area around Nineveh (Bagg 2000: 169-224). 

276 Developments in Farming, Animal Husbandry, and Technology 

The construction of a canal which ran from the river Khosr near the city of 
Kisiri, some 16 kilometers away from Nineveh, was the first step in Sennacherib's 
irrigation program. Together with a royal park, this canal is mentioned for the 
first time in 702 BC. Grape vines, fruit and olive trees, spice plants, and cypresses 
grew in this park. Also connected with this phase of the project was the granting 
of land to the inhabitants of Nineveh for the establishment of orchards north of 
the city. These were watered by secondary canals and ditches. The main function 
of the large canal was therefore the irrigation of the orchards above the town. 
Remains of a canal which approached the town from a northeasterly direction 
were discovered in the 19th century. 

The next step in the development of Nineveh's irrigation network was the 
construction of the Mount Musri canal system, from C.694BC. During the seven 
years which separate these projects the city-scape changed considerably. In 699 BC 
an artificial swamp was created which served for the regulation of high water in 
the canal during the spring. Reeds from this marsh were used as building material 
in the palaces. Moreover, two further gardens were established to the north of 
the city, in the same area where an aqueduct was built. Spring water from Mount 
Musri (Jabal Bashiqa, about 20 kilometers away from Nineveh) was led into 
reservoirs and by means of canals and/or canalized waits to the Khosr. How the 
water was then brought to Nineveh is unknown, because no remains of this canal 
system have yet been discovered. The Mount Musri canal system was constructed 
to irrigate orchards and grain-fields to the south of the city in the summer 

The northern canal system was a combination of natural and artificial water- 
courses, by which the fields that lay to the north of the town, between the cities 
of Tarbisu (modern Sharif- Khan) and Nineveh, were watered, allowing cereals 
and sesame to grow there. This canal system probably represents the third step 
in Sennacherib's hydraulic program and was undertaken between 694 and 691 BC. 
Three stretches of the northern canal system have been identified in the field and 
the existence of a fourth has been postulated (Oates 1968: 50-51; Reade 1978: 
158-65; Ur 2005: 325-35). The system collected the water of several wadis, 
namely the Rubar Dahuk, the Wadi Bahandawaya, and the Wadi al-Milah. In 
connection with the canal works, rock reliefs were carved at Maltai, Faida, and 
Shiru Malikta. The last stretch of this canal system, from the juncture of the Wadi 
al-Milah with the Tigris to the city, ran parallel to the river and reached Nineveh 
from the northwest. 

The Khinis canal system, built C.690BC, was the last stage in Sennacherib's 
irrigation program. The Gomal river, which rises in the Kurdish mountains, was 
dammed near the village of Khinis (about 50 kilometers northeast of Nineveh) 
and brought to a tributary of the Khosr by means of a main 35 kilometer-long 
canal known as "Sennacherib's canal." The canal head, a masterpiece of Assyrian 
hydraulic engineering, comprised a dam, intake works, and a 300 meter-long 
canal, with stone parapets and a tunnel. From this tunnel the water was conducted 



via a rock-cut channel. The canal head was located in a gorge to the north of 
Khinis, where the remains of these works were found together with gigantic rock 
reliefs and inscriptions. Halfway down to Nineveh, near the village of Jerwan, it 
was necessary to build an aqueduct for the canal to cross a valley. The aqueduct 
was 280 meters long, 16 meters wide and 7 meters high (9 meters, including the 
parapets). It was supported by five corbeled arches (Jacobsen and Lloyd 1935). 
Some stone blocks were carved with inscriptions, in which the king appears as 
the builder of the aqueduct and the canal. This was the only hydraulic engineer- 
ing project which could have supplied the area north of Nineveh as well as the 
southern area with irrigation water. 

In connection with the first step in Sennacherib's irrigation works a park 
related to the palace was mentioned above. On a wall relief from AssurbanipaFs 
North Palace at Nineveh a hilly park appears (Figure 14.3). On top of a hill 
planted with broad-leafed trees and conifers stands a pavilion and a stele on which 
the Icing is depicted. A canal, fed by an aqueduct, flows from right to left. Its 
corbeled arches closely resemble those of the Jerwan aqueduct. Many secondary 
canals or ditches branch off from the feeder canal. This scene probably shows 
one of Sennacherib's parks, which was fed with irrigation water by means of an 
aqueduct built in or near Nineveh (Bagg 2000: 196-8). 

Figure 14.3 Irrigated park with aqueduct (7th century BC) at Nineveh, North Palace 
(BM 124039) (drawing by the author). 

278 Developments in Farming, Animal Husbandry, and Technology 


For the environmental conditions in the Ancient Near East see Butzer (2000); for the 
regime of the Tigris and the Euphrates see Ionides (1937). The standard work about 
agriculture in Iraq is Wirth (1962). Booher (1974) offers a very clear explanation of 
surface irrigation techniques. Interesting studies concerning irrigation agriculture in Syria 
are collected in Geyer (1990a). A concise history of ancient Near Eastern irrigation 
according to the written sources can be found in (Bagg 2003). Excellent studies on the 
written sources for Mesopotamian irrigation can be found in BSA 4 (1988) and 5 (1990). 
A study of the available sources on water lifting devices in the ancient Near East is offered 
in Bagg (2001). 

For the earliest irrigation in Mesopotamia, see Oates (1969) and Helbaek (1972). An 
accurate summary of irrigation in southern Mesopotamia can be found in Postgate (1992: 
173-90). A brief English discussion of irrigation agriculture at Mari is offered in Lafont 
(2000); for the excavator's description of the canal remains at Mari, see Margueron (2000: 
68-82). For a short and precise summary of the Khabur canals and agriculture in the 
region, see Morandi Bonacossi 1996: 95-101, 194-204), and Ergenzinger and Kiihne 
(1991). For a comprehensive study of Assyrian irrigation works, see Bagg (2000). For 
hydraulic works probably related to irrigation in Urartu (Lake Van area), see Garbrecht 
(2004). For the cuneiform sources concerning water rights, see Bagg (2000: 63-72).